Anthropic says threat groups linked to Russia and China abused Claude AI for cyberattacks, espionage, scams, malware development, and other malicious operations. The company disrupted the activity, banned the accounts, and strengthened its safeguards after tracking abuse tied to ShinyHunters, Midnight Blizzard, and GTG-10007. #ShinyHunters #MidnightBlizzard #GTG-10007 #Claude
Keypoints
- Anthropic found multiple threat groups misusing Claude for harmful cyber and influence operations.
- ShinyHunters used AI-assisted pipelines to harvest credentials and accelerate breaches.
- Midnight Blizzard used Claude to support malware development, phishing, persistence, and exfiltration.
- GTG-10007 used Claude to orchestrate espionage, vulnerability research, and exploit development.
- Anthropic blocked the actors, updated guardrails, and notified authorities, partners, and victims.